Is this amp picking up Morse, and can you read it ?

[GrimJosef]

I have a stereo amp in for work at the moment which is significantly noisier in one channel than the other. At least some of the noise is associated with poor layout of the signal ground wiring, which I will set about trying to fix tomorrow. Listening to it this afternoon I thought I could hear Morse among the general hash being picked up. Can anyone tell me whether it really is Morse, or some other form of coded signal ? If it is Morse can you pick out what's being sent ? Apologies in advance for the very poor recording quality. The sound was faint and the general background noise in the workshop, including this evening's strong wind, is all too audible.

Cheers,

GJ

[Graham G3ZVT]

In short, no I don't think it's Morse, neither machine or hand sent.

Trying to read it nonetheless I get
MOE T E T E --..- E T [...]
Perhaps --..- is an obscure punctuation mark, but nor one I have ever used.

All moot because I don't think it is Morse!

Is there anything "digital" in the amp, a strobed digital display for example

Or perhaps it's an embedded processor in a kitchen appliance, perhaps you could investigate with a portable radio?

I have to say though I can see why you thought it might be Morse code, it does have that 3:1 ratio that Morse has.

[GrimJosef]

Thanks for the help. I'd like to quieten the noisy channel down to the level of the other one in any case - I was just curious about the noise's possible source.

There's nothing digital in the amp (it's a heavyweight 20-odd year old stereo job with a pair of 211 output valves running single-ended and a few ECC8x's doing the small-signal gain and the driving). There's a cordless phone, Siemens AL14H, sitting in its cradle in the same room. I can still hear the 'Morse' when everything else is switched off. The level doesn't change if I power the amp from a synthetic 240VAC supply (UPS running on batteries) isolated from the mains and with, or without, the mains earth connected. Nor does it change if the amp's baseplate is fitted or absent.

Cheers,

GJ

[paulsherwin]

Amps can sometimes pick up SW, particularly on the phono input, but I've no experience of it happening with valve kit. I guess it could happen if the conditions are right.

[joebog1]

Switch off your mobile phone!!!!

Sounds EXACTLY as my hand made amp if I have mobile phone in my pocket.

Joe

[Radio Wrangler]

Pull the plug and listen to it as the amp fades. If it stops suddenly, it's mains-borne grot.

David

[GrimJosef]

I did try that. It seems to fade with the amp's HT voltage, although that is very fast. The output valves are directly heated triodes, so they switch off quickly anyway, and the very high HT (1250V on the reservoir, 1100V on the smoother) means the capacitors aren't high value. Since the amp is single-ended the quiescent current draining the caps is large too. So I tried the synthetic 240VAC power supply which is separate from the mains. The noise was still there. I am suspecting the cordless phone.

Cheers,

GJ

[David G4EBT]

I'd agree with the analysis of Graham, Rambo1152, as to the letters at the beginning: MOETETE, but it's very slow - maybe 5WPM, so would be unlikely to be machine-sent code, and I can't think why anyone would want to send Morse manually at such a slow speed. As others have said, I think it's much more likely to be mains-borne electromagnetic interference.

[Radio Wrangler]

Another source of clues is the time of day it happens at. Does it happen all day or have hours on and off?

A few miles away from me there was a Naval radio station. Quite powerful, everything metallic in the vicinity throbbed to MTO MTO MTO and a listing of reply frequencies. It was amazing where Dah-dah, dah, dah-dah-dah turned up!

THe transmission was clean enough, but the strength just overloaded everything.

David

[Gi4CZW Cliff]

hello GJ seems to be repetitive. Have you a longer recording ? Cliff.

[Radio Wrangler]

The pull-the-plug test seems to indicate it's being picked up off-air, maybe by speaker leads acting as the antenna.

The sound is rather 'fizzy' something spiky with lots of harmonics in the audio. But that could be an artefact of unintentional rectification. Real CW would just come over as thump-thumps with no audio tone, so something else is going on.

Having a scan around with a wide range receiver or spectrum analyser might find the real signal. It ought to be pretty strong to be noticed

tried sticking 10 foot of wire into a scope?

What sort of amplifier is it?

David

[eaglework]

Sounds very much like the interference you can get when a mobile phone is syncing with its base station. Used to be nightmare is radio studios until people realised it wasn't a good idea to put a mobile on the desk near un-terminated input channels.
So I suspect its something to do with the cordless phone.

[GrimJosef]

Thanks for the continued interest folks.

The amp is a 20-odd year old Gamma Rhythm - looks like this https://i.pinimg.com/originals/83/c2...7fbbccc8fc.jpg (that's a stock picture, not the one I'm working on).

Practicallly all of the electronics (not the HT supply) are on a single pcb which also supports the small-signal valve sockets. These amps could be a bit untidy underneath and the grounding arrangement is unusual. This one doesn't seem to be original either. The output stage ground, but not the 211s' grid leak resistors, are connected via a common 100R lifter to the chassis. IMHO the lifter is far too weedy to handle a serious fault current. The mains safety earth goes direct to the chassis (good). The small-signal stages, including the 211s' grid leaks, have separate L and R grounds on the pcb and these are returned via the phono stage's input leads' screens to the amp's input sockets (all their grounds are commoned, with the L and R channels kept separate) and then to the chassis via a 3R3 lifter - one for each channel. The phono stage pcb ground is not isolated from the other small-signal stages, so to avoid loops the screens of all the line-level input leads and of the phono stage output lead are unconnected at the input selector switch. Likewise the screened leads from the switch to the volume control, and from the volume control back to the pcb, are only grounded at the pcb end, not the switch end.

The pcb is not L-R symmetric nor is it side-by-side duplicated. It's a mix of the two, and this is why one channel picks up more noise than the other. On the noisy channel the signal ground from the volume control goes round the houses before getting back to the line stage input and thence to the phono stage input. Part of it is shared in common with a couple of HT smoothing cap grounds, which might explain an excess of 100Hz. I can only partially null this with the 211 filament balancing pot, so there must be some phase-shift between the two 100Hz contributions.

I'm afraid I did some lash-up re-routing of the signal ground last night which lowered the noise, including essentially eliminating the 'Morse', and this morning I've wired that tidily, so I can't make any more recordings. You're right though Cliff - it did sound repetitive, but the repeats were short. There were at least a couple in the recording I made.

The 'Morse' was there for more than an hour last night David, but I didn't check for it this morning. Didcot is about as far as you can get from the sea in England, so I don't think it would have been the Navy . The Army have some presence in these parts though. And they fly helicopters out of Benson a lot.

Cheers,

GJ

[dodgy-dxer]

Not Morse, not even close

As others have said its more like Mobile phone pickup

or some other data signal

Fred

[Radio Wrangler]

I normally give boutique amplifiers a wide berth. I don't have the right religion to appreciate them. But I don't see a thermionic rectifier sticking up prominently in its parade of glassware, so I assume there are silicon HT rectifier diodes hidden carefully out of sight, and I assume these are the only semiconductors in the equipment.

If all the signal stages are thermionic then the usual unintentional demodulators aren't present and valves usually take a lot more signal to get up enough non-linearity to do the job. I'm beginning to suspect something mains-borne. Do you get mains fluctuations in time with the sounds, I wonder. Might be worth looking at neutral-earth as well. It doesn't seem to be simple Morse, because some of the dahs seem substantially longer than the others. (US railway Morse used three different lengths.... but they're long gone and even further away than the sea)

I'm wondering about some kind of on/off controller of something on the mains. Have you ever heard the motor of an old Kenwood Chef mixer running on part-throttle..... very hit and miss.

Single ended big valves.... but the transformer box seems a bit mean-sized for the power those anodes suggest. Could things be a bit close to saturation?. Is there any Morse in the transformer hum?

I always get nervous when I come across peculiar theories about grounding.

Oh, a last thought, RF rectification could be happening in the output stages of whatever's driving it.

There'll be a conspiracy theorist along soon suggesting it's strange life-forms (not as Jim Kirk knows it) created and living in the Diamond Light Source, trying to make contact, asking to be let out

David

[Graham G3ZVT]

As you mentioned a DECT phone in the room can we assume you have eliminated the base unit and handset as the cause, and any mobile phones too?

Do you have pair of those Power-Line Adapters in use? The things that extend a wired Ethernet connection via the mains wiring and/or create a second WiFi access point at the distant end?

[knobtwiddler]

I presume you have measured and compared frequency response, harmonic distortion and noise floor between channels? It does sound like the asymmetrical layout is the issue as you suggest, but if you find one channel starts to roll off by 30K, whereas the other starts to rise, you could have a clue right there. Does a 10K square look the same L/R?

[GrimJosef]

David, boutique amplifiers can be a trial. But when they act up their owners do seem to want them to be sorted out. Someone in the trade told me I'm developing a reputation for being prepared to go in where no-one else will.

The HT (1250V, 120mA or so) is indeed solid-state (bridge) rectified. So are the DC supplies to the 211 filaments (the 50Hz hum would be intolerable otherwise) and the small-signal valves also run off DC heaters. Finally I am running with an anti-parallel high-current diode ground-lifter too. But all these semiconductors are 'just' diodes.

It wasn't coming down the mains lead. It was there, and just as loud, when I ran off a battery-powered supply.

Graham, there were no mobiles (very) nearby. Both mine and my wife's are usually switched off. The ground rewiring has eliminated the problem now (or it's simply stopped) so I can't check whether disconnecting the DECT phone makes it go away. Which is a shame. I strongly suspect that was it, but now I can't confirm it. I've got no WiFi extenders in use. My basic WiFi box (Virgin Superhub) is not in the workshop, although its signal is usable there.

Knobtwiddler, I haven't run the measurements yet. The R channel, with the dots and dashes (or whatever they are), does have more noise on it. Bits of everything really - continuous broadband with a rising profile below a few hundred Hz, 100Hz and harmonics from the various rectified supplies, a bit of 50Hz, that regular few Hz ticking that you might just be able to hear on the 'Morse' recording. I fear that drawing useful conclusions about the audible noise under quiescent conditions from measurements mad when the amp is being driven would be a challenge though (for me, at least ). I would say though that apart from the dots and dashes the R is not hugely noisier than the L.

Cheers,

GJ

[Radio Wrangler]

The diodes lifting the ground could be a demodulating non-linearity. Changing internal ground wiring affecting things seems to play this way too. But why one channel and not t'other?

A jolly good mystery.

David

[knobtwiddler]

But why one channel and not t'other?

Indeed. Has anyone suggested a grid-stopper R that's gone out of whack? Or a BW limiting cap? Analysis of the good channel against the bad one would be highly revealing of a fault of this nature.

At the risk of appearing over-simplistic, I have to say that, if you have access to a real-time FFT analyser and dummy load, merely looking at the noise floor and inputting a 1K sine to see harmonics typically reveals 90% of faults. Many scopes offer an FFT function (which may not be good enough to really look at the noise floor profile) - and there are also freeware options which will work with soundcard. Pro analysers such as the Prism DScope or RTX6001 are getting cheaper all the time. Simply looking at the angle of roll-off in terms of dB p/octave can diagnose faults. If the good channel has a nice gaussian noise distribution, whereas bad channel appears 'filtered' in some way, you have a powerful clue.